PPS trigger VS system time Trigger

[Loic VINET]

Hello

I am using "digital discovery" connected to a PC or a raspberry.

I have a trigger given by a GPS having 100 ms up and 900 ms down.

I run the here-include python program (see scriptLV) at a frequency of 5 seconds, and I get the trigger time using the function 'dwf.FDwfDigitalInStatusTime(hdwf, byref(psecUtc), byref(ptick), byref(pticksPerSecond))'.

I should have the same time at a millisecond precision, but I have a variation at each iteration and I don't understand why.

        Time                    |  ms 
------------------------------------
2023-12-12 18:52:58 | 432.076782
2023-12-12 18:53:08 | 432.460197
2023-12-12 18:53:18 | 432.157247
2023-12-12 18:53:28 | 433.663812
2023-12-12 18:53:38 | 436.344122
2023-12-12 18:53:48 | 437.429318
2023-12-12 18:53:58 | 438.901731
2023-12-12 18:54:08 | 440.093428
2023-12-12 18:54:18 | 442.817301
2023-12-12 18:54:28 | 443.583707

What is the explanation? Is it the internal clock of the PC or the Raspberry?

Regards

scriptLV.py

[attila]

Hi @Loic VINET

For Digital Discovery, Analog Discovery 3, ADP3X50 and newer devices, on software connection the computer time is synchronized with the device.
The time stamp for each capture is given by the device timer. The trigger time resolution is given by sampling rate (8/10ns) and for Digital Discovery at 100MHz or higher is 1.25ns
The accuracy of the device timer is given by the device's crystal oscillator, but AD3, ADP3X50... can also use external reference clock.

See the following example: DigitalIn_TriggerTime.py

With 1Hz pulses generated by the same device the trigger time is precisely 1s step:

By triggering on pulses from another device you can see the frequency difference between crystal oscillators, here ~4us/sec: